1. Field of the Invention
This invention relates generally to devices for projecting pictures onto large viewing screens; and more particularly to such devices that project laser beams via reflective liquid-crystal light valves to form such pictures. The invention has its most important applications in such projection of moving pictures.
2. Related Art
a) Known potential of lasers--Since the advent of the laser, people have been trying to find new ways to use lasers in projecting pictures of one kind or another, for large audiences. This is both natural and reasonable, since lasers offer several important characteristics that are relevant in large-image projection.
As will be seen from the following recap of these characteristics, one would expect these characteristics to be responsible for a predominance of laser projection systems in large-screen displays for both video home use and theater-scale displays. Indeed, several powerful large international companies have attempted--at monumental cost--to develop such equipment for market.
Therefore, while reading the following discussion of laser advantages for large-screen projection sources, please bear in mind this overriding question--why are large-screen laser projectors not common in the marketplace?
(i) energy efficiency--All other things being equal, the amount of light needed to show any kind of picture on a projection medium (viewing surface) is proportional to the area to be covered by the picture. Optical energy is therefore of utmost importance in a large-format projection system, and it is necessary to pay for visible optical energy with electrical energy.
In such transactions it is well understood that some conversion inefficiency is unavoidably involved as a sort of tariff--in other words, that a sizable fraction of the electrical energy used will go into invisible forms of energy such as heat, or near-infrared and ultraviolet radiation. Normally there is relatively little objection to this price in itself, but the question does arise of just how sizable a fraction one can afford.
With nonlaser light sources, this concern is compounded when taking into account the additional surcharge for optical energy that is visible but goes off in directions other than into the collecting optics of a projector. Most nonlaser sources (incandescent hot-filament or arc lamps) radiate approximately equally in all directions. The amount of visible light that can be directly collected from such a source into an optical system is typically less than a tenth of the visible light produced.
It can be dismaying to pay for many times the amount of electrical energy used--even that which is directly used to make visible light, setting aside consideration of the conversion efficiency discussed above. Therefore it is common to provide reflectors behind the source, or more generally speaking to try to surround the source with reflectors to help capture a greater geometrical fraction of the visible energy. Such efforts, however, complicate and compound the management of heat thrown off due to those same conversion inefficiencies considered above.
A laser, though of course itself a costly article, greatly improves all these energy economics. Since its optical emissions are directional, essentially all the emitted light can be very easily captured for use.
Furthermore, to a significant extent the spectral components an be controlled so that minimal energy is wasted in infrared or ultraviolet radiation. A laser is therefore far more energy efficient than other sources--with respect to both raw conversion efficiency of electricity into visible light and geometrical capture of that visible light.
Lasers and their power supplies do give off heat, and this must be managed. In comparison with a typical arc lamp or like device, a laser is vastly more favorable with respect to the amount of heat, the temperature involved, and the difficulty of collection.
(ii) brightness--With most types of light sources, increasing the amount of light available calls for fabrication of a source that is scaled up in all three dimensions, more or less equally, and therefore greatly complicates the process of collecting the light and drawing off heat.
To make a brighter laser, it is necessary in essence to make a laser which is just like one that has various desirable known properties, except with a bigger tube. Over a small range of brightness increases, furthermore, what is needed is only a longer tube. Heat management with a longer tube reduces to using the same hardware, but more of it, as with a shorter tube. Even if brightness requirements do call for increased diameter too, the elongated character of most laser structures tends to distribute and thus mitigate the problems of power and heat management.
With a bigger laser, all the greater amount of optical flux can be made to go in essentially the same direction and into essentially the same projection system as the corresponding smaller laser. These oversimplifications of course slight some practical considerations such as design of power-supplies, cooling, and lasing modes, but summarize an important way--for purposes of image formation--in which lasers differ from other light sources.
(iii) contrast--Several properties of lasers tend to enhance contrast in a projected image. The simplest of these is once again inherent directionality, which facilitates both collection of input illumination and handling of an image, with minimum crosstalk between different portions of the beam or the image.
Contrast is enhanced by avoiding such crosstalk--or in other words preventing the spill of a cast over an entire image frame, from bright image areas. Such undesired spill corrupts areas that should be dark. Further enhancement of just the same sort arises from the inherent collimation of a laser beam.
Equally or more important, when most modern image-modulation devices are taken into account, is the inherent monochromaticity of a laser beam. Other sources emit light over the entire visible spectrum, requiring subdivision into spectral segments, and physical separation into distinct beams that can be separately modulated and then recombined to give full-color images.
In either type of system, laser or nonlaser, the final optical stage--i.e., the projection lens--is preferably broadband since it preferably carries all the colors in a common beam; for this purpose a high-quality achromat is desired. The benefits under discussion apply to all earlier optical stages, where the functions being performed are much fussier and complicated than the final projection stage.
With such other sources, each distinct beam carrying a separated spectral segment is already broadband, either complicating or degrading the effectiveness of all optical effects or manipulations. These include everything from perturbation of simple focusing (chromatic aberration) to the operation of sophisticated image-modulating devices (see below).
Since operation of lenses, polarizers, prisms, dichroics and image modulators are all wavelength-dependent, the operation of virtually all optical components in a projection system using such other sources tends to scatter light away from the precise bright-region positions where it should be. The result is to create a kind of halo about such positions--or, again, depending on the brightness contours of a particular image, even to produce a filmy bright cast over much of a scene that should be darker.
Also of great importance is the inherent polarization of a laser beam. Many large-screen projectors of the present day employ an image-writing stage that controls a high-intensity light beam by spatial modulation of the beam.
As discussed more fully in later sections of this document, almost all such modulators rely upon formation of a latent image in polarization state (or as it is sometimes called, a "phase object"). This image is later developed by passage through some form of polarization analyzer.
Other projectors intensity-modulate a scanning spot of a high-intensity beam; here too, the phenomenon most commonly exploited to accomplish modulation is the polarization of the beam. For all such applications a laser is ideally suited, first of all because no light need be discarded (or recaptured through a complicated optical train) merely because its polarization state does not match what can be used by a modulator.
More significant to contrast enhancement is the relative sharpness (i.e., narrowness of angular range) of laser polarization, in comparison with polarization obtained through a common polarizer. Because of this, in areas of a latent image that should be bright (calling for passage of a beam through the downstream polarizer), the polarization state provided when using a laser source is defined more sharply; the same is true for areas that should be bright and call for extinction. The latent image therefore is potentially brighter where it should be bright, and darker where it should be dark--or, in other words, has better contrast.
It is true that the latent image yet remains to be developed through a polarizer, leading to some imprecision in isolating for projection the polarization state that is nominally correct. Nevertheless--even based upon the sharper polarization definition in the latent image alone--both the beam passage in bright areas and the beam extinction in dark areas are better.
(iv) sharpness--Another benefit due to the inherent collimation of a laser beam is that it produces sharper images. This is partly associated with contrast enhancement, due to the wavelength- and polarization-dependent effects discussed above. In a scanning-spot projection system (whether amplitude-modulated or not), laser-formed images are sharper also in part because of the capability of a highly collimated beam to be focused to a fine spot.
In image-modulation systems, laser beams are able to traverse great distances without degradation of spatial modulation. In other words, a spatially modulated beam can carry an image over a long distance without becoming blurry. For a laser system, this performance characteristic may be more associated with favorable divergence properties than with collimation.
In any event, except for contrast effects already discussed, the capability of a good arc-lamp-based projector to produce a sharp image at a distance may be about as good as a laser-based system heretofore--provided that the image is projected onto a screen or other viewing surface that is:
(1) flat or very gently curved, PA1 (2) essentially at right angles to the beam, and PA1 (3) not moved toward or away from the projector after the projector is set to produce a sharp image. PA1 e) Speckle--This well-known term describes a now-familiar phenomenon of laser illumination, a coarse and very bright granular pattern of light that shimmers with tiny movements of the viewer's eyes. Speckle is highly undesirable in image projectors for displaying ordinary pictures (movies, television shows etc.) because it pervades the images and distracts from the informational or dramatic content of the show. PA1 (1) use of such reds with acoustooptic modulators (AOMs), or PA1 (2) direct, electrical amplitude modulation of the source lasers. PA1 a vector, bitmap or other computer file scanned from an image or generated in a computer, or PA1 amplitude-modulated laser-diode illumination swept two-dimensionally across the control stage, or PA1 images from a small transmissive liquid-crystal display modulator, in turn written by signals not derived from traditional broadcast video signals, or PA1 other entire frames without interlace, or PA1 a raster whose lines cross a short dimension of a picture frame, or PA1 motion-picture film color separations, or PA1 a still image from a slide or overhead-projection transparency, or a color separation made from such a slide or transparency, or PA1 a live image optically coupled, without electronic intermediary, to the control stage. PA1 means for reshaping the initially circular laser-light beam to a shallow, wide laser-light beam, and PA1 means for scanning the shallow, wide laser-light beam over the screen. PA1 an interior of a dome, or other building having internal surfaces that are not generally normal to a projection direction, PA1 an exterior of a dome, sculpture, monument, or other structure having external surfaces that are not generally normal to a projection direction, PA1 a waterfall, PA1 a water fountain, PA1 fog or a cloud, PA1 ice, PA1 a scrim in front of a curtain or screen, PA1 a plurality of scrims in optical series, PA1 one or more trees, PA1 grass, vines or other foliage, PA1 a hillside or other landscape, or other receding surface, or PA1 an array of people or other animals or other discrete objects, or combinations thereof, at diverse distances from the projecting means.
In other words the prior-art laser projector may have little advantage in sharpness as such if the projection medium is all at the same distance from the projector, and there is an opportunity to adjust the projector for the actual projection distance.
Cases in which these conditions fail are discussed in the following paragraph. In both types of systems, laser and nonlaser, the ability to maintain image sharpness as such over long distances depends to a major extent on the quality and size of the final projection lens.
(v) infinite depth of sharpness--Laser systems have a unique and major advantage over white-light systems, in projection onto projection media that are at varying distances from the projector. Such media also can be positioned or oriented so that they are not all at a common, preset projection distance.
These media can include, for example, surfaces that are strongly angled to the projection beam. In the vector-graphics part of the laser-projector field, this is a well-known characteristic--which I have sometimes termed "infinite focus". It is also possible with other types of laser-transmitted images, including both vector- and raster-scanned spots as well as images projected with spatial modulators.
It has been suggested to me that the term "infinite focus" is a misnomer, in that "focus" refers to formation of an image at a preset "focal plane" (sometimes in the retina) by convergent light rays from various parts of a lens system. Such convergence requires adjustment of the optical system for a specific projection distance--a process with which of course nearly everyone is familiar. My phrase "infinite focus" derives from the concept of "depth of focus", combined with the idea that laser-transmitted images seem to have infinite depth of focus.
As I now understand, however, laser beams when used to project images in such a way as to obtain this effect are not focused at all. The image is not formed by convergence of rays from different parts of a lens, either at a preset "focal plane" or otherwise.
Rather an image can be impressed upon a laser beam by so-called "spatial modulation" of the beam. This means that each pencil of rays from the laser carries a specific, fixed part (e.g. pixel) of the image. Laser beams are initially collimated so that the ray pencils are all parallel, never crossing one another or converging.
It is possible to force a laser beam to converge to a rather fine point (of course only an approximation of a point) by interposition of a lens that does focus all the rays. For present purposes it would not make sense to do this, since there would be no image--only a single bright spot--and indeed this is never done in a system that displays the "infinite" effect.
Instead the spatially modulated beam is simply directed to a viewing medium, where the ray pencils are stopped and so make the impressed image visible to viewers. In practice such a beam can be expanded, to form a large image on a large viewing medium, and for this purpose a substantially conventional lens may be employed--and within the constraints of pixel or raster-line visibility the image will be sharp but never "focused".
The other half of the phrase "infinite focus" is also somewhat inaccurate since there are some limits to the depth, along the projection direction, at which images appear sharp. These limits are imposed by beam divergence and other diffraction effects.
For reasons that will appear after I have introduced my invention, however, these effects should never come into play in a laser-based system properly designed and assembled for image projection. Therefore in other parts of this document I have replaced my earlier terminology with the phrase "infinite depth of sharpness" or simply infinite sharpness.
It has been recognized that this deep-sharpness effect is of potentially great value for special effects. Some of this value has been actually achieved in some vector systems, as will be explained below, but the much greater potential for raster images has not been realized in practice heretofore.
The importance of the previously posed question as to the nonappearance of laser projectors in the marketplace should now be apparent. Some reasons for that peculiarity will appear from the following sections of this document.
b) Vector-scanning laser systems--Generally speaking this term refers to free-form movement of laser beams from any point on a projection medium to any other point, and following any specified trajectory (e.g. curved) rather than a preset framewise pattern as discussed in the next section.
(i) light-show style--Historically these were the first displays for large audiences, and are straightforward to produce since equipment was minimal and artistic opportunities maximal. In most cases the beams are neither amplitude modulated nor focused (a small-diameter laser tube yielded a small spot for entertainment purposes), and a relatively slow sweep is usually employed so that audiences can perceive the spot motion itself as well as the trajectory. Since the color effects of the independent laser beams are an important part of the show, there is no point in forming or sweeping a combined beam.
I mention these early systems primarily because--as long as the beams are not focused--a primitive sort of equivalent of infinite sharpness is enjoyed for each beam independently. That is to say, beams can be projected onto surfaces at considerably varying distances from the lasers without changing spot size.
The beneficial uses of this phenomenon are entirely familiar to designers and operators of these shows as a sort of special-effects trick that can be used to enhance light-show imagery. The desirability of extending this phenomenon to infinite sharpness as related to projection of whole picture images is accordingly also believed to be known in this field.
It will be understood, however, that in the light-show context spot size does change, to the extent that the beams are spread out on a viewing surface that is angled to the projection beam. (Depending on audience position--i.e., whether the audience is looking essentially along the projection direction or along a normal to the surface, or from some other direction--the stretching may not be visible in its entirety, or at all.)
It is very important to recognize this sort of spreading on an angled viewing surface, and to distinguish it from failure of the beams to be sharp. An analogous spreading/sharpness distinction arises later in discussing whole-picture-image projection.
(ii) graphics--As in the now-familiar vector graphics of computer programs such as CAD/CAM, Corel@ Draw, Visiog and so forth, the use of vector graphics in a laser-based projection system is well understood and highly versatile. It may be used to provide economically and quickly a simple, static production nameplate, or a more elaborate moving display for similar purposes, or of course cartoons for entertainment etc.
In this case the beams may be amplitude modulated for more complex effects, and the beams may be combined into a composite beam that is swept as a unit--in which case the entire resulting image may enjoy infinite sharpness provided that the beams are simply projected and not focused. As will be seen, vector-graphics projection is of only secondary interest for present purposes.
c) Raster-scanning systems--The topic now turns to reproduction of whole picture images that are generalized, in the sense that the projection system is a neutral vehicle for display of any raster-based image. The projection-system raster can be set to match traditional or conventional broadcast television, whether U.S. interlaced or otherwise, or to match a high-definition television format--or to match a conventional computer-monitor format, or any other well-defined raster specification.
(i) amplitude-modulated spot, with separately swept beams--In essence such a system would be a direct laser-projector analog of a conventional television set, requiring amplitude modulation at video speeds, and for each color independent two-dimensional sweep. Such devices may never have been put into practice, but they are well represented in U.S. Pat. No. 3,524,011 of Korpel (1968), assigned to Zenith Radio Corporation. (Korpel's independently swept beams optionally share a common projection lens.)
Such a system cannot provide accurate infinite sharpness of a full-color image, as introduced above, since Korpel's separately swept individual-color beams emanate from spaced-apart points (possibly even spaced-apart projection lenses) and can therefore accurately converge to form a registered image only at a preselected plane. If, however, the projection distance (or audience distance) is kept much larger than the spacing between the origination points or lenses, and the inherently collimated beams are not focused, registration error at differing projection distances can be made negligible and a semblance of infinite sharpness can be obtained.
(ii) amplitude-modulated spot with sweeping of a combined beam--A device of this type should have true infinite sharpness, since what is swept is a unitary beam (again provided that the system does not bring the beam to a sharp focus). Systems with this type of configuration and particularly employing solid-state lasers are disclosed by Knize, noted earlier, as well as U.S. Pat. No. 5,534,950 to David E. Hargis and U.S. Pat. No. 5,614,961 to Frank C. Gibeau, Ph. D. Amplitude modulation in these systems is by electrical control of the lasers.
It appears that these systems may have considerable promise, but are not to be found in the marketplace. It would seem that for these devices with present-day available components the laser power at certain needed wavelengths, or the modulation response speed, or the overall economics, or combinations of these considerations, are inadequate for realistic commercial exploitation.
d) Line-scanning systems--The great bulk of reported and patented developments in laser projectors is of this type, using a separate acoustooptic modulator (AOM) for each primary color. A seminal patent in the linewise AOM regime is U.S. Pat. No. 3,818,129 of Yamamoto, assigned to Hitachi. In such a system each AOM is a crystal driven by an acoustic wave propagating laterally (with respect to the laser-beam path) and modulated by one video raster line at a time.
The compressions and rarefactions of this input modulation in the AOM create or write a phase-retardation pattern within the crystal, extending transversely from one side of the crystal to the other and representing optical modulation in one primary color for an entire video raster line. In the most-advanced forms of these systems, just as the formation of this retardation pattern is completed a laser is pulsed to provide a light beam intersecting the pattern at right angles.
This reading-beam pulse length is very short compared with the propagation speed of the acoustic wave through the crystal, so that in effect the laser illumination is able to stop the motion of the raster line. The laser beam in effect reads the entire retardation pattern, and upon leaving the crystal has impressed upon it--in phase retardation--a latent image of the entire raster line.
This image is then developed, as suggested earlier, by a polarization analyzer or equivalent, downstream of the crystal. The result is an image of one primary color component of the raster line, which is then preferably combined with like images for the other two primaries, formed in separate AOMs.
At some point in the optical system, whether before or after the modulation stage, each of the three individual primary-color laser beams or the composite beam must be shaped to form a wide, shallow beam cross-section. For reasonable optical efficiency within the modulators it appears preferable to use a more-common beam aspect ratio in passage through the modulators--i.e., to perform the shaping after the beam has passed through the modulators, though before the final projection lens. Considerable variation in such aspects of the design, however, is possible.
The composite beam is enlarged and projected to a particular position vertically on a viewing screen, forming a three-color raster line for viewing by the audience. The process is repeated for successive lines--but shifting the vertical position progressively down the screen--to construct an entire image frame, and then for subsequent frames to produce moving pictures.
The vertical position for each raster line is controlled by a rotating polygon or other vertical-sweep device, so that successive lines are displaced to successive appropriate positions on the screen. This sweep, it is important to note, follows the modulators--i.e. is introduced downstream, along the optical path, from the modulators--as exemplified, for instance, by U.S. Pat. No. 5,255,082 of Tamada, assigned to Sony.
Thus in AOM systems the slot-shaped beam is scanned or stepped only on the projection screen, not on the modulators. Though capable of moderately high contrast (over 300:1 in certain military projectors), high resolution, reasonably good color saturation, and infinite sharpness, this type of system is subject to important limitations and also certain qualifications as explained below.
It appears that some of the largest and most sophisticated corporate participants in the laser-projector race have persistently placed their money--many millions of dollars of it, over many years--on the acoustooptic modulator entries. These include Sony, Schneider, TRW and IBM as well as a host of lesser players.
For all that wagering, none of the AOM entries is seen to place or even show, today. Many have dropped out entirely.
As suggested near the beginning of this "background" section, resources invested in laser projectors have been wholly disproportionate to performance. The question remains why this pattern continues.
(i) light inefficiency and energy loss--This is the dispositive consideration for AOM-based systems. Unfortunately the compromises that enable achievement of the favorable parameters listed above also reduce, to an unacceptably low level, the light efficiency of the modulators and the system in general. The only laser projectors built in this way that actually operated to produce excellent image quality were military systems that required extremely large, high-power, expensive lasers.
(ii) low bandwidth--Another element that suffers in these systems is the capability to follow rapid action in a scene. This may be related to persistence (or propagation speed) in the AOM crystal, or the modulation constraints that follow inherently from the need to refrain from outpacing the pulsed-laser optical reading system.
(iii) complex optics--Many optical stages are needed in an AOM system. The military projectors mentioned above, though they operated continuously for two years and always maintained certain military specifications of brightness, resolution and contrast, had more than forty-five optical elements. Each optical surface attenuates undesirably.
(iv) stepped, slot-shaped beam--The special significance of these features will be seen in later portions of this document. For purposes of the present "background" section, it suffices to point out that use of this type of beam is required by, and directly associated with the nature of the line-at-a-time modulator:
Since the modulator processes one raster line at a time, the pulsed beam on which this modulation is impressed must necessarily correspond in shape to the wide, shallow aspect ratio of one raster line. It would not be possible to operate a one-raster-line-at-a-time modulating system with any other beam shape.
Similarly, it follows necessarily from the generation of a complete raster line in optical form that the optical system must include an optical stepper or continuous scanner of some sort--to shift the target position successively down the viewing screen for the successive raster lines, as described earlier. Even a continuous scanner, in this type of system, amounts to a stepper since the beam is pulsed only intermittently, once per raster line. It would not be possible to operate a one-raster-line-at-a-time modulating system without some sort of stepper.
To the best of my knowledge it has not been reported in the prior art that a slot-shaped beam, or a stepping system for such a beam, might confer any other benefits upon a laser projection system.
Now before going on from vector-, raster-, and line-scanning (AOM) systems to take up systems that employ some very different kinds of modulation, I shall pause and digress to discuss some very important special considerations peculiar to laser operation. As will be seen later, these are matters of particular relevance to my invention.
It has been explained to me that speckle is an interference pattern formed within the eye. Although in principle present with other sources too, speckle is not ordinarily visible with such sources. Those skilled in the art recognize that the speckle effect can be made negligible by introducing various kinds of either phase confusion or relative motion, as between the laser source and the eye.
Heretofore, however, actual equipment called into service for accomplishing this has fallen far short of the elegant. Many elaborate schemes of greater or lesser cost and complexity are described in the literature.
One such "speckle eliminator", which is among the more complex but demonstrates the seriousness of the problem, is presented by Hargis, mentioned previously. Hargis introduces several approaches, "each of which introduces an optical path randomizing [medium] at an intermediate . . . plane within the projection optics".
One of his systems is "a spinning diffusion plate" which works at "transverse plate velocities in excess of a few centimeters per second" but suffers from "transmission inefficiency (.about.50%) . . . large numerical aperture . . . and . . . general bulkiness." Transmission is improved "to the 85%-95% regime" by substituting "a thin sheet of wax supported between glass plates."
Another system is a "flowing fluid diffuser" using "a highly turbid fluid", suffering from "low transmission efficiency with the inconvenience of a pump and associated plumbing." A third, relying not on flow but on "Brownian motion", Hargis rejects because "its transmission efficiency is limited, compared with what presently appears to be the best available system described below".
His favored choice is a "novel nutating plate" which "takes advantage of the desirable properties of wax laminate diffusers". It involves a screen--
"supported on springs, and caused to vibrate in a plane . . . perpendicular to the projection axis of the video image beam . . . by orthogonal electromagnets . . .
"Motion relative to two orthogonal axes is induced in plate 25, together with a 90-degree phase shift between those motions, in order to avoid periodic moments of zero velocity which would be associated with simple harmonic motion along a single axis. The result is a non-rotating diffuser which undergoes rapid nutation, much in the manner of the contact surface of a[n] orbital sander. Hence, all regions of the image are subjected to the same motion. An excursion of 1 millimeter at 60 Hz provides constant transverse velocity of about 20 cm sec.sup.-1. This yields an inexpensive device which is barely larger in cross section than the imaging beam itself."
Provision of his illustrated device, plus a system of electromagnets and associated electrical drive, may not be expensive but it is certainly elaborate and surely diffuses--and thus randomly redirects and wastes--expensive laser energy.
Other workers have proposed a great variety of systems (likewise severely overcomplicated, in most cases) for elimination of speckle. Representative are U.S. Pat. No. 5,272,473 teaching a transducer that generates surface acoustic waves in a projection screen, U.S. Pat. No. 5,506,597 proposing an array of mirror cells movable between two positions in conjunction with a magnifying element, U.S. Pat. No. 5,274,494 disclosing use of a Raman cell to introduce optical sidebands, U.S. Pat. No. 5,233,460 counseling division of laser light into three separate beams and introducing differential delay or polarization rotation before recombination, U.S. Pat. No. 3,633,999 similarly advising a splitter to make many separate beams whose speckle patterns mutually cancel, and U.S. Pat. No. 4,511,220 describing two polarizing beam splitters and a totally reflecting right angle prism that form a composite beam with mutually incoherent components.
Very generally speaking, speckle elimination systems of which I am aware exhibit two common drawbacks. They add otherwise unnecessary mechanical or electromechanical equipment, and more importantly they subtract light.
f) Gamut and saturation--patents and other technical literature that touch on the selection of wavelengths for the primary colors in laser projectors, by and large, have favored color conventions or standards approaching those of commercial broadcast television. The most important of these conventional wisdoms relates to selection of wavelengths for use as the primary red.
It is well known that wavelengths close to the visible-color chromaticity envelope provide the broadest and best base for building a capability to display rich, saturated colors. Nevertheless leading workers in the laser-projector field have taught away from use of a long-wavelength red.
For example, U.S. Pat. No. 5,255,082 of Tamada, assigned to Sony, strongly rejects use of laser lines in the region of 647 nm for a primary red beam. Tamada offers the reasoning that such wavelengths should be avoided because they are weak in the spectra of certain lasers which he prefers.
Following suit is U.S. Pat. No. 5,136,426 of Linden, assigned to Advanced Laser Projection. Linden warns that the--
"red light component produced by the krypton ion laser requires four-to-five times the power as the comparable power of an [argon] ion laser . . . . The krypton red light component is at a wavelength that the human eye is not as sensitive to and therefore makes it difficult to balance with the other colors to give a complete color scale with reasonable power.
"The [argon] ion laser in combination with a dye laser is therefore preferred . . . . The dye laser preferably converts light energy of a shorter wavelength to a longer, tuneable wavelength."
Like other leaders in this field, Tamada and Linden counsel use of wavelengths in the range of 610 nm for primary red, generally based on rationales such as presented above.
It appears that one underlying motivation for such a choice may have stemmed from the use of commercial video standards or conventions--NTSC,
or HDTV--which consistently favor the 610 nm range. This historical choice, in turn, appears to have arisen not truly because of apparent luminosity to the human eye but rather from the limited availability of television-display phosphors during early color video development.
Another interesting historical development in the laser projector field is the prevalent technique of filtering out certain cyan lines that are present in popular lasers--particularly argon lasers, which are a good choice for providing both blue and green lines. There seems to be a high likelihood that the cyan light is discarded because it prevents ready mixing of accurately neutral colors (black, white and gray), as well as ideal rendition of all other colors--when 610 nm lines are chosen for the red primary.
The choice of laser light at 610 nm for red thus has complicated repercussions--particularly since the cyan light in an argon-laser beam amounts to some forty percent of the total power or energy in the beam. Discarding that large fraction of the beam power is a profligate waste, when a major challenge in the laser projector field is finding enough energy at a reasonable price to form an adequately bright large image.
Whether because television phosphors lacked the capacity for deeper red or because of their need for greater brightness, present laser-projector workers stress the NTSC-based luminance chart and the 610 nm red options--and thus forsake the broader color gamut available in both film and computer monitors, as well as the ample beam power readily available in the cyan lines.
Some writings in the laser-projector field, such as the Tamada and Linden patents, do at least mention the possibility of longer-wavelength primary reds. All such writings are limited to either:
As will be seen, neither of these paths is part of the genealogy of my invention.
g) Laser types proposed or used--It is well known, at least in concept, to employ lasers of a great number of different types for laser projectors. In particular it is known to employ gas, dye and solid-state lasers in this field.
(i) gas--Many subtypes are known, but foremost in this category are argon lasers for spectral-line groups in the blue and green, and krypton lasers for red. Thus argon gas laser beams are commonly split for separate modulation in separate AOMs that receive blue and green image-data components, while a krypton gas laser beam is modulated in a third AOM that receives red image-data components.
These lasers are relatively straightforward to operate and adjust. They require neither pumping nor tuning. They require neither mixing nor frequency-doubling. Accordingly they provide good efficiency as to both electrooptical energy and human-operator efforts.
(ii) dye--In the opinions of many workers in this field, dye lasers are of particularly great value because they are tunable (particularly to 610 nm). In the opinion of this writer, reliance on tunability is a handicap because of the extra operator attention which it demands, as well as the high cost of tunable mirrors and other needed paraphernalia.
Dye lasers are considerably less user-friendly than gas lasers, on account of their requirements for management of an additional pumping stage at the front end and mixing stage at the back. In most cases they also consume profligate amounts of extra energy in generating light at frequencies that are not wanted but merely needed for purposes of subtraction or addition to obtain desired frequencies.
This waste may be acceptable in high-end consumer or boardroom equipment, where literally conspicuous consumption can be a virtue. It is highly questionable, however, in a cost-conscious commercial environment, for example a light-hungry projector system for driving a monumental IMAXO-style screen or an outdoor-spectacle system which projects images onto, actually, monuments and other structures.
(iii) solid state--These devices may in the end become the only sources that make economic sense, for use in my invention as well as other types of systems. At the time of writing, reasonable sources are available in red and green.
No adequate solid-state laser exists, however, for use as a blue primary source in even a large consumer/boardroom unit. Solid-state blue lasers adequate for use in large outdoor displays would appear to be at least some years in the future.
It is true that for such special applications a very large number of individual very small solid-state lasers can be ganged to amass a mighty beam. The overall economics (and possibly ancillary procedures) of that approach appears unfavorable relative to the present invention.
h) Liquid-crystal "device" modulators--Unlike the AOM, a liquid-crystal "device" or "display" (LCD) modulator provides modulation over an entire frame. Here it is possible to flood an entire frame at a time, and project the resulting full frame to a projection screen or other viewing medium.
(i) some leading work in the field--Active current effort on advanced LCD modulators that operate on unpolarized beams is seen from researchers at Kent State University (see SMPTE Journal, April 1997). Earlier LCD efforts correspond to U.S. Pat. No. 5,040,877 of Blinc, assigned to Kent State; U.S. Pat. No. 5,517,263 of Minich, assigned to Proxima Corporation; U.S. Pat. No. 4,851,918 and U.S. Pat. No. 4,720,747 of Crowley, assigned to Corporation for Laser Optics Research; and also U.S. Pat. No. 5,485,225 of Deter, assigned to Schneider.
(ii) visible electrode structure--All LCD modulators are operated in transmission. That is to say, in such a system a laser beam is projected completely through the entire device from one side to the other.
All these devices accordingly require direct electronic writing of the desired image electronically rather than optically--and this in turn requires one or another form of multiple-electrode structure, in a pattern that is spread over the entire frame. These electrodes are nominally transparent, and indeed are not readily visible in displays of modest size, such as for instance less than five feet along a diagonal.
In theater-size and larger formats, however, the electrode edges are quite conspicuous. These patterns are distracting and intrusive, leaving LCD modulation essentially unusable for high-quality imaging in theater and outdoor applications, unless all of the audience is at a very great distance from the screen or other projection medium.
(iii) no infinite sharpness--Also a drawback for such large-scale applications is the fact that these LCD units fail to preserve the laser property, described earlier, of maintaining sharp imaging at widely varying projection distances. Various special-effects potentialities are thereby foreclosed.
i) Liquid-crystal light valves--These liquid-crystal light valves (LCLVs) are to be carefully distinguished from the liquid-crystal display or device modulators discussed just above. Whereas an LCD operates in transmission and requires passing the projection beam through electrodes in the image-writing (input) stage of the modulator, an LCLV operates in reflection and has entirely separate image-writing and projection stages.
The image-writing stage may have electrodes, or may be written optically or thermally, but all such activity is entirely isolated from the projection stage by an opaque mirror. There is one, unitary electrode in the projection stage but its edges are ordinarily outside the image frame.
(i) development of the LCLV--Pioneering work with LCLVs is due entirely to Hughes Aircraft Company and Hughes-JVC Technology Corporation. This is seen in a series of patents extensively elaborating the LCLV and its usage in many variants over two decades. These include U.S. Pat. No. 4,019,807 of Boswell (1977), U.S. Pat. No. 4,127,322 of Jacobson, U.S. Pat. Nos. 4,343,535 and 4,378,955 of Bleha, U.S. Pat. No. 4,425,028 of Gagnon, U.S. Pat. No. 5,071,209 of Chang, U.S. Pat. No. 5,363,222 of Ledebuhr, U.S. Pat. No. 5,398,082 of Henderson, U.S. Pat. No. 5,428,467 of Schmidt, U.S. Pat. No. 5,450,219 of Gold, and U.S. Pat. No. 5,465,174 of Sprotbery (1995).
A particularly important precursor of the LCLV is attributed to Dr. Bleha. Particularly helpful expositions of the working principles of these ingenious modulators appear in the Boswell and Jacobson patents. Apparently an LCLV may be a twisted-nematic type, a birefringent type, a hybrid of the two, etc.
(ii) structure and operation of an LCLV--Common to the several LCLV variants is a basic laminar configuration in which an input or writing stage first develops a voltage that varies spatially within the device frame, in accordance with brightness variations that constitute an image to be projected. An output or reading stage has a polarization-influencing characteristic--such as a particular index of refraction, corresponding to a particular optical phase delay.
The writing stage and reading stage are separated by an opaque mirror, and the whole assemblage is sandwiched between two transparent planar electrodes. By virtue of these electrodes, voltages developed in the writing stage are applied to the reading stage.
The spatially varying voltage induces corresponding spatial variations in the polarization-influencing characteristic of the reading stage. Meanwhile polarized light--the reading beam--is introduced into the output or reading stage, reflected from the internal mirror and returned toward the projection screen.
The spatial variation in index causes the desired image-brightness variations to be expressed as a spatially varying polarization field, carried by the light beam leaving the reading stage. As described earlier, this polarization field is decoded or developed by a polarization analyzer so that the beam carries a spatially varying intensity field, which is perceptible to the eye as an image. For color images, this strategy is replicated for each of three primary colors.
The resulting beam or beams are projected (with or without combination into a common projection beam) in a substantially conventional way through a projection lens to a viewing medium. Whereas the writing stage may be excited with very low-intensity light as for instance from a small CRT (or by low voltages applied to an electrode matrix, or in other ways), the reading stage is preferably excited with extremely intense, projection-level illumination--such as, in the Hughes work, a high-current arc lamp.
Evidently Hughes personnel have explored the use of LCLVs with, exclusively, such incandescent sources ("white" light). One reference, however, does propose the use of LCLVs with laser sources--and that is not a Hughes document but rather is the above-noted patent of Minich (Proxima). Both types of usage are discussed below.
(iii) image projection using incandescent-lamp sources--Regardless of other optical conditions, broad-spectrum conventional light sources cannot provide the infinite-sharpness characteristic. It goes without saying that the Hughes projectors, operated as described in all the Hughes patents, necessarily operate by actually focusing images on a projection screen, with the associated shallow depth of focus. Accordingly these systems are incapable of the earlier-mentioned special-effects applications that rely on infinite sharpness.
(iv) full-frame--Most of the Hughes patents describe operation with the high-intensity "reading" or output stage of the LCLV modulator flooded continuously by projection light, or in other words all illuminated at once. This type of operation offers a particularly appealing simplicity and elegance: in essence the entire projection frame is opened and held open, for whatever input may be written to the input stage.
The output for regions of the frame that are not being written, however, is simply dark. Thus for instance if a very small but bright pen-light type of flashlight could be pointed onto the writing stage and played about manually, presumably a mammoth searchlight would appear to be--in real time--correspondingly wandering about on the projection medium, which might be for instance the exterior of a very large building. Subject to contrast limitations, the projection medium would be substantially dark in regions corresponding to writing-stage regions not illuminated by the pen-light.
(v) poor energy economics, and brightness nonuniformity--The full-frame LCLV Hughes system is, however, subject to several drawbacks. First, per the above introductory subsection on laser vs. nonlaser comparison, as in most other projection systems the light from an incandescent source is emitted in essentially all directions. Only a small fraction of this omnidirectional radiation can be effectively captured for guiding into the LCLV, and the remainder becomes a source of heat-management problems.
Second, the light collected from a high-intensity source is typically nonuniform across the frame in which that light is collected. This too can be mitigated, and in conventional ways including use of frosted (i.e., diffusing) elements--but such solutions further scatter optical energy with only limited directionality, and so inevitably further aggravate the already unfavorable collection geometry. Special lensing, too, may be used to reduce central bright spots, but at yet-additional cost--both monetary and thermal.
Third, most writing stages operate incrementally--in other words, based upon some sort of scanning input such as a raster-driven or vector-graphics-driven spot of light, which inherently can be active in only a very small portion of the writing-stage frame at any given moment. The costly or even precious high-power light beam, however, is directed indiscriminately to the entire frame, including mostly unreceptive regions that are not being written at any given moment. This mismatch of written and read regions is mitigated by the persistence characteristic of the LCLV--that is, the continuing capability of a written region to pass reading light, for a length of time perhaps equal to a tenth to a fourth of the period of an entire frame, after the writing in that region stops. Thus the unfavorable factor is not on the order of thousands, only on the order of four or ten--but still distinctly unfavorable.
Fourth, yet more energy loss is incurred in beam masking to fit the image shape & projection frame. Whereas collection systems typically yield beams that are circular, projection frames are square or (particularly for widescreen movies) rectangular.
In the case of masking down a circular beam 11 (FIG. 26) to a square projection format 774, for example, the discarded chordal areas 775 amount to about thirty-six percent of the area of the circle--as is verified by simple arithmetic later in this document. Thus 36% of the energy in a circular beam is wasted in masking to a square frame.
Worse, in masking to a three-by-four screen format 874 (FIG. 27) the discarded fractions 875 come to 39%. In going to the popular widescreen nine-by-sixteen format 974 (FIG. 28) the lost fractions 975 are nearly 46%, close to half of the optical energy in the circular beam.
(vi) polarization analyzer--Now turning from energy losses somewhat in the direction of performance, the intrinsic contrast ratio of an LCLV although high is far from perfect, particularly since polarization extinction for broad-spectral-band light is hard to control. (As noted previously, the operation of polarizing devices is wavelength-dependent.)
Thus a perceptible glow may pass through an LCLV to the projection medium in regions that should (based on the written image) be dead black. In this way some of the costly optical energy extracted from the omnidirectional source--and still remaining after the several inefficient processes discussed above--is used to illuminate areas that are dark in the desired image.
(vii) vertically swept "slot"--Several of the Hughes patents are direct testament to the intractable character of these problems. The above-mentioned Henderson, Schmidt and Gold patents in particular lay out these same difficulties and discuss a proposed solution.
Henderson teaches simply shaping of a white-light beam, from an incandescent source, into a shallow slot-shaped beam--and scanning that beam across an LCLV modulator. In this case, since the light source itself is continuously operating, a continuous sweep produces a continuum of overlapping successive beam positions rather than a discrete-stepping effect. Henderson's goal is to greatly improve energy uniformity, masking, read/write efficiency and contrast of an LCLV system by placing the reading light in precisely the region where the writing is taking place.
Evidently, as it appears, Henderson was not wholly successful in this--since the companion Schmidt patent explains at column 2 (lines 48 through 56), and also at column 9 (lines 30 and 46) that Henderson's approach, considered alone, suffers severely from the loss of "telecentric behavior" of the optical system, and also from chromatic aberration. Schmidt notes that the purpose of his own invention is to restore "telecentric behavior" and mitigate adverse chromatic effects.
A telecentric optical system is defined in the Gold patent as a system in which all "chief rays" are made to parallel the optical axis of the system. A chief ray, in turn, is by definition a ray that originates at an off-axis point of an object or source and crosses the axis. Like chromatic aberration, these are characteristics of conventional white-light systems in which, for example, rays from various points of an object which extends transverse to the axis are collected in a lens and redirected--many typically crossing the axis--to construct an image also transverse to the axis (but located at another point along the axis).
Schmidt proposes resolving the Henderson problems through particular forms of rotating polygonal deflectors that are transparent, and ingeniously configured to preserve telecentricity. Gold teaches use of a more conventional reflective rotating polygon, but coupled with somewhat elaborate optical elements to pre- and postcondition the slot-shaped beam for deflection at the polygon--also to preserve (or restore) telecentricity.
Despite these yeoman efforts, it appears that Hughes has never used the scanning-slot system commercially. Not even the most-recently introduced projector models--or technical papers--from the Hughes development group suggest any movement toward adoption of the Henderson/Schmidt/Gold system.
Perhaps this is due to the difficulty of forming a white-light source beam into a very shallow, very wide slot-shaped beam, without discarding so much light that the overall system is unacceptably inefficient and impractical. Henderson, for example, mentions (column 6, line 55) that brightness in at least the vertical cross-section of the beam is Gaussian, and suggests masking off a substantial portion of even that cross-section to avoid using the skirts of the Gaussian beam. In any event it seems that the scanning-slot beam--if not simply inoperative--was a dead-end side trip, in the course of developments at the birthplace of the liquid-crystal light valve.
(viii) image projection using laser sources--The previously mentioned Minich patent proposes to use LCLVs with laser sources--including red laser lines in the neighborhood of 620 nm. Minich asserts that his LCLV-based apparatus is "substantially similar . . . to the system [using a transmissive LCD modulator], except that the [LCLV] apparatus operates reflectively rather than transmissively."
By lumping these devices together somewhat indiscriminately, Minich suggests less than full appreciation for their major differences. As mentioned earlier, the transmissive LCD devices are objectionable for very-large-format projection because of conspicuous electrode patterns which they display.
Neither the problems of beam-shape matching and contrast nor the possibilities of scanning slot-shaped beams are taken up by Minich--in either his above-noted patent or his more-recent one, U.S. Pat. No. 5,700,076. These problems are just as important with laser sources as with the Hughes white lamps.
Likewise the problem of speckle in systems using LCLV modulators is never taken up by Minich in those patents. It is substantially impossible to operate a laser/LCLV projector without addressing this obstacle.
Minich furthermore fails to address the desirability of infinite sharpness, although this represents a major application for laser projectors. The conventional understanding is that the image-forming mechanisms of LCLV modulators destroy laser-beam coherence and thereby foreclose achievement of infinite sharpness.
Still further, in the patents mentioned above Minich says nothing of the problems of brightness uniformity. Whereas beam nonuniformity in white-light LCLV systems is significant, in a laser-beam LCLV system it is of the utmost importance--because laser beams are subject to a number of artifacts that become plainly visible on the projection screen if a laser beam is simply expanded to flood an LCLV reading stage.
To fill in certain portions of his disclosure, Minich refers to documents of Texas Instruments Incorporated (column 5, line 58) and of Hughes (column 9, line 58). The overall focus of the Proxima development program, as suggested in the Minich patent, is upon very compact, lightweight and inexpensive projectors that are very unlike the very large, high-quality Hughes product (and two orders of magnitude lower in price). Actual Proxima machines on the market appear to correspond to the more-recently issued '076 Minich patent mentioned above, not to anything in Minich '263.
All in all, it appears that the disclosures in the '263 Minich patent are conceptual rather than practical. It may offer, as the foregoing enumeration of omissions suggests, a less than completely enabling disclosure.
j) Marketplace considerations--The foregoing discussion indicates some answers to the question posed earlier, "why are large-screen laser projectors not common in the marketplace?" The answer is that numerous practical problems attendant the real-world design and manufacture of a commercially viable laser projector have not been answered.
One device that might provide a key to solution of some of these problems--the liquid-crystal light valve--has not been associated with laser projectors either in the market-place or (notwithstanding the Minich '263 patent) in any meaningful, practically oriented enabling publication. No product or publication has revealed how to provide infinite sharpness, or otherwise how to project images on irregular projection surfaces having dramatically varying projection distances.
No teaching in the art has revealed how to defeat speckle without adding elaborate equipment appendages that subtract light. The art has never resolved, in marketplace terms, the problems of brightness, contrast, energy efficiency, masking, or illumination of nonwriting regions which Henderson, Schmidt and Gold attempted to address.
As can now be seen, the related art remains subject to significant problems, and the efforts outlined above--although praiseworthy--have left room for considerable refinement.